Unit Economics for Infrastructure Companies

Most infrastructure founders can tell you their gross margin. Few can tell you their fully-loaded cost per customer. This gap is where profitability goes to die.

Unit economics in infrastructure are different from SaaS or consumer businesses. Your costs are variable. Your customers are technical. Your value is invisible until something breaks. Traditional metrics don't capture the full picture.

The Full Cost Stack

When founders calculate unit economics, they typically count direct costs: infrastructure, payment processing, maybe some support. But infrastructure businesses have hidden costs that compound at scale.

Layer 1: Direct Infrastructure Costs

These are easy to measure. They show up on your AWS or GCP bill. Most founders stop here. That's a mistake.

Layer 2: Operational Costs

These costs scale with customer count but aren't always attributed per customer. A monitoring system that costs $5,000/month serves all customers, but the cost per customer changes as you scale.

Layer 3: Hidden Costs

These are the costs that surprise founders. They don't scale linearly. They compound. A fintech company might spend 10% of revenue on fraud prevention. A data platform might spend 15% on compliance. These costs are real, recurring, and often underestimated.

Target Margins by Category

Not all infrastructure businesses should target the same margins. Your category determines your cost structure, which determines your target margins.

If you're below these ranges, you have a structural problem. Either your costs are too high, or your pricing is too low. Both are fixable, but you need to know which problem you have.

Calculating True Unit Economics

Here's the framework we use to evaluate infrastructure companies. It's more comprehensive than standard gross margin calculations, but it's also more accurate.

Step 1: Define Your Unit

What's the atomic unit of value delivery? For a payment API, it's a transaction. For a data platform, it's a query or GB processed. For an authentication service, it's an active user. Define this clearly.

Step 2: Calculate Direct Costs Per Unit

Step 3: Add Operational Costs

Take your monthly operational costs (monitoring, security, support infrastructure) and divide by monthly transaction volume. This gives you operational cost per unit.

Note: This cost decreases as you scale, which is why infrastructure businesses improve margins over time.

Step 4: Add Hidden Costs

Fraud, compliance, and support costs are harder to measure but critical to include. Use historical data or industry benchmarks.

Step 5: Calculate Fully-Loaded Unit Economics

This is your true unit economics. Not the 97% you get from subtracting only direct costs. The 71.7% you get from including everything.

How Unit Economics Change at Scale

Infrastructure businesses have interesting scaling dynamics. Some costs decrease per unit as you scale. Others increase. Understanding this is critical for projecting profitability.

Costs That Decrease at Scale

Costs That Increase at Scale

Common Unit Economics Mistakes

Mistake 1: Ignoring Fraud Costs

Fintech founders often model 1-2% fraud rates based on early data. Real fraud rates at scale: 5-15%. This difference destroys unit economics. Model conservatively.

Mistake 2: Underestimating Support Costs

Early customers are forgiving. They'll debug issues themselves. Enterprise customers expect white-glove support. Budget 10-20% of revenue for support at scale.

Mistake 3: Not Accounting for Failed Transactions

You pay infrastructure costs for failed transactions too. If 10% of transactions fail, your effective cost per successful transaction is 10% higher than you think.

Mistake 4: Assuming Linear Scaling

Infrastructure costs don't scale linearly. You hit inflection points where you need to rebuild systems. Budget for these architectural upgrades.

Using Unit Economics for Decision Making

Unit economics aren't just for investors. They're a decision-making framework for operators.

Pricing Decisions

Know your fully-loaded cost per unit. Price at least 3x that for sustainable margins. If you can't charge 3x your costs, you have a business model problem, not a pricing problem.

Customer Segmentation

Not all customers have the same unit economics. Enterprise customers might have higher support costs but lower fraud rates. Small businesses might have higher fraud but lower support needs. Segment and price accordingly.

Feature Prioritization

Features that improve unit economics (reduce fraud, automate support, optimize infrastructure) should be prioritized over features that just drive growth. Sustainable growth requires healthy unit economics.

Fundraising Strategy

If your unit economics are strong, you can be capital efficient. If they're weak, you'll burn cash at scale. Fix unit economics before scaling aggressively.

The Bottom Line

Unit economics are the foundation of infrastructure businesses. Get them right, and you can build a sustainable, profitable company. Get them wrong, and you'll burn cash at scale.

Calculate your fully-loaded costs. Include everything. Model conservatively. Price accordingly. Monitor continuously. Improve systematically.

This discipline separates infrastructure companies that scale profitably from those that burn cash at scale.

If you're building infrastructure and want help modeling your unit economics, let's talk.